Informality, Welfare and Productivity

46
Informality, Welfare and Productivity Carmen Pagés Inter-American Development Bank Labor Markets Unit (LMK)

description

Informality, Welfare and Productivity. Carmen Pag és Inter-American Development Bank Labor Markets Unit (LMK). How does informality affect workers’ welfare?. I will argue that informality affects welfare not in the way we (I?) thought… - PowerPoint PPT Presentation

Transcript of Informality, Welfare and Productivity

Page 1: Informality,  Welfare and Productivity

Informality, Welfare and Productivity

Carmen PagésInter-American Development Bank

Labor Markets Unit (LMK)

Page 2: Informality,  Welfare and Productivity

How does informality affect workers’ welfare?

• I will argue that informality affects welfare not in the way we (I?) thought…

• That is…not by condemning informal workers to “bad” jobs… (at least not to all)

• But rather, lowering aggregate productivity and with it, the earnings of most workers (formal and informal, as well as firm’ owners)

• And by.. making labor and social policy extremely difficult and ineffective

Page 3: Informality,  Welfare and Productivity

This presentation:

1. How does informality affect individuals´ welfare? – Are formal workers better off? Which ones?

2. What is the effect of informality on aggregate productivity?

3. How does informality affect labor and social protection policies?

4. Policies to increase formality • Credit and formality

Page 4: Informality,  Welfare and Productivity

The one-slide presentation

1. Informal workers are not necessarily worse off (particularly the self-employed with low education attainment)

2. Yet Informality creates important aggregate problems:

• It kills the efficacy of labor and social protecction policies as we know them;

• It reduces aggregate productivity.

3. A more proactive approach needed:

• Re-designing labor and social policies to be informality-proof• Or tackling informality: reducing cost of formality, and

Increasing its benefits.

• More access to credit helps.

Page 5: Informality,  Welfare and Productivity

Part I: Informality and Individual welfare:Based on

“Is Informality a good Measure of Job Quality?:

Evidence from Job Satisfaction Data”Lucia Madrigal, IADBCarmen Pagés, IADB

Page 6: Informality,  Welfare and Productivity

Informality: Choice or/and exclusion?

• Wage differentials –positive but affected by selection bias.– If corrected using semi-parametric methods or

panel data much smaller differentials.– Wages do not necessarily reflect welfare

• Mobility studies– if workers value formality they should voluntarily move to formality and involuntarily away from formality. – “revealed preferences” – Problem. In most cases we do not observe if

movements are voluntary or not.

Page 7: Informality,  Welfare and Productivity

In our work

• Use job satisfaction data to assess whether informal jobs are less valued than formal jobs

• If workers in formal jobs enjoy rents they should report higher levels of job satisfaction than informal workers.

Page 8: Informality,  Welfare and Productivity

Job satisfaction: the literature

• In similar approaches, subjective job satisfaction measures have been used to understand the – Determinants of job quality (Clark, 2004; Sousa

Poza and Sousa Poza, 2000) – Determinants of utility of unemployed workers (is

unemployment voluntary?) – Whether self-employment enjoy rents.

To data these studies conducted mostly in developed countries.

Page 9: Informality,  Welfare and Productivity

Rather than adhering to any particular definition of informality, we distinguish between:

• Self-employed;• Employed in firms of more than 10

employees with benefits;• Employed in firms of more than 10

employees without benefits• Employed in firms with less than 10

employees without benefits;

Page 10: Informality,  Welfare and Productivity

We examine the determinants of Job satisfaction as a function of:

• X1: Observable worker characteristics (gender, age, educ, health status, civil status)

• Z1: Job category cum benefit variables:• Self-employed;• Employed in a small firm• Employed in a large firm without benefits• Employed in a large firm with benefits

• Z2: Objective job characteristics (earnings, industry, occupation, hours )

• Z3: Subjective job characteristics (well remunerated, opportunities for promotion, job is stressful, job is dangerous, monotonous, good work schedule, job is insecure, )

Page 11: Informality,  Welfare and Productivity

Methodological issues (I)

• Omitted variable bias: Correlation between job satisfaction and subjective variables may be driven by innate unobservable individual traits (i.e optimism).

– We control for different degree of optimism as in van Praag (08)• We make use of individuals’ valuations (Vi) about other aspects

(health policy, education policy, transportation)/– Regress different Vi on Xi;– Obtain principal components of residuals. – Add the first principal component (Ki) in (1).

• Sample selection issues, particularly for women, may be important –need to control for it.

Page 12: Informality,  Welfare and Productivity

Methodological issues (II)

• Further omitted variable bias issues: Even after controlling for optimism one could argue that the relationship between type of job and job satisfaction is driven by unobservable variables.

– For example, less able individuals choose informality and at the same time have lower expectatives and therefore higher JS.

• Need panel data to properly account for that, however data contains information on whether workers prefer working as self-employed or salaried.

Page 13: Informality,  Welfare and Productivity

Data (I)

• Data for three low income countries (Guatemala, Honduras, El Salvador)

• Collected by Statistical National Offices in 2007

• Nationally representative samples (18-64 years old)– Honduras: 8288 individuals.– El Salvador: 1082 individuals– Guatemala: 1400 individuals

• One individual randomly selected within the household answers a longer questionnaire

• In practice some over-representation of women

• Re-weighting

Page 14: Informality,  Welfare and Productivity

Data (II)

• Data contains info on: – Individual (including health status) and

objective work characteristics– Job satisfaction and subjective

appreciations of job characteristics

• Whether workers would prefer being salaried or self-employed– Job satisfaction defined as “are you

satisfied with the job you do?”

Page 15: Informality,  Welfare and Productivity

Job Characteristics by Job CategoryWeighted Sample of Workers 18 to 64 years old-Honduras

Self-Employed

Salaried_small

Salaried large-no benefits

Salaried large-

benefitsDependent Variable

Job Satisfaction 0.865 0.731 0.837 0.943Objective Job Characteristics (Z 2i )

Hours (monthly) 149.77 165.10 181.32 163.07Log Earnings (monthly) 7.632 7.469 8.432 9.023Primary activities 0.379 0.344 0.086 0.014Manufacturing 0.137 0.074 0.327 0.080

Construction, retail, trasnsport and storage 0.399 0.362 0.178 0.135Utilities, financial and personal serv. 0.085 0.219 0.409 0.771

Subjective Job Characteristics (Z 3i )

Stressful 0.503 0.483 0.534 0.646Dangerous 0.327 0.353 0.349 0.239Monotonous 0.167 0.188 0.146 0.075Progress Opportunity 0.644 0.514 0.718 0.767Good Schedule 0.852 0.818 0.823 0.872Well Remunerated 0.583 0.506 0.570 0.557Insecure Job 0.106 0.291 0.197 0.086

Ranking does not correspond with traditional distinctions based on firm size

Page 16: Informality,  Welfare and Productivity

Job Characteristics by Job CategoryWeighted Sample of Workers 18 to 64 years old-Honduras

Self-Employed

Salaried_small

Salaried large-no benefits

Salaried large-

benefitsDependent Variable

Job Satisfaction 0.865 0.731 0.837 0.943Objective Job Characteristics (Z 2i )

Hours (monthly) 149.77 165.10 181.32 163.07Log Earnings (monthly) 7.632 7.469 8.432 9.023Primary activities 0.379 0.344 0.086 0.014Manufacturing 0.137 0.074 0.327 0.080

Construction, retail, trasnsport and storage 0.399 0.362 0.178 0.135Utilities, financial and personal serv. 0.085 0.219 0.409 0.771

Subjective Job Characteristics (Z 3i )

Stressful 0.503 0.483 0.534 0.646Dangerous 0.327 0.353 0.349 0.239Monotonous 0.167 0.188 0.146 0.075Progress Opportunity 0.644 0.514 0.718 0.767Good Schedule 0.852 0.818 0.823 0.872Well Remunerated 0.583 0.506 0.570 0.557Insecure Job 0.106 0.291 0.197 0.086

Job characteristics

Page 17: Informality,  Welfare and Productivity

Job Characteristics by Job CategoryWeighted Sample of Workers 18 to 64 years old-Honduras

Self-Employed

Salaried_small

Salaried large-no benefits

Salaried large-

benefitsDependent Variable

Job Satisfaction 0.865 0.731 0.837 0.943Objective Job Characteristics (Z 2i )

Hours (monthly) 149.77 165.10 181.32 163.07Log Earnings (monthly) 7.632 7.469 8.432 9.023Primary activities 0.379 0.344 0.086 0.014Manufacturing 0.137 0.074 0.327 0.080

Construction, retail, trasnsport and storage 0.399 0.362 0.178 0.135Utilities, financial and personal serv. 0.085 0.219 0.409 0.771

Subjective Job Characteristics (Z 3i )

Stressful 0.503 0.483 0.534 0.646Dangerous 0.327 0.353 0.349 0.239Monotonous 0.167 0.188 0.146 0.075Progress Opportunity 0.644 0.514 0.718 0.767Good Schedule 0.852 0.818 0.823 0.872Well Remunerated 0.583 0.506 0.570 0.557Insecure Job 0.106 0.291 0.197 0.086

Page 18: Informality,  Welfare and Productivity

Job Characteristics by Job CategoryWeighted Sample of Workers 18 to 64 years old-Honduras

Self-Employed

Salaried_small

Salaried large-no benefits

Salaried large-

benefitsDependent Variable

Job Satisfaction 0.865 0.731 0.837 0.943Objective Job Characteristics (Z 2i )

Hours (monthly) 149.77 165.10 181.32 163.07Log Earnings (monthly) 7.632 7.469 8.432 9.023Primary activities 0.379 0.344 0.086 0.014Manufacturing 0.137 0.074 0.327 0.080

Construction, retail, trasnsport and storage 0.399 0.362 0.178 0.135Utilities, financial and personal serv. 0.085 0.219 0.409 0.771

Subjective Job Characteristics (Z 3i )

Stressful 0.503 0.483 0.534 0.646Dangerous 0.327 0.353 0.349 0.239Monotonous 0.167 0.188 0.146 0.075Progress Opportunity 0.644 0.514 0.718 0.767Good Schedule 0.852 0.818 0.823 0.872Well Remunerated 0.583 0.506 0.570 0.557Insecure Job 0.106 0.291 0.197 0.086

Page 19: Informality,  Welfare and Productivity
Page 20: Informality,  Welfare and Productivity

Table 3Probit-Mg Effects All Workers-Honduras 18-64 years oldDependent Variable: Job Satisfaction

(1) (2) (3) (4) (5) (6) (7) (8)Salaried_small -0.1257*** -0.1226*** -0.1267*** -0.0755*** -0.0658*** -0.0398* -0.0629*** -0.0384*

(0,0205) (0,0218) (0,0227) (0,0238) (0,0217) (0,0229) (0,0216) (0,0228)Salaried_lg_nobenefits -0,0165 -0.0411** -0,0301 -0.0460** -0,0082 -0,0073 -0,0071 -0,0065

(0,0152) (0,0176) (0,0187) (0,0218) (0,0182) (0,0211) (0,0181) (0,0209)Salaried_lg_benefits 0.0961*** 0,0575 0.0757** 0,0541 0.0633** 0.0617* 0.0625** 0.0607*

(0,0260) (0,0362) (0,0330) (0,0428) (0,0291) (0,0344) (0,0288) (0,0342)Other Controls None Ind. charact. Ind, objective. Ind, objective. Ind, objective. Ind, objective. Ind, objective. Ind, objective.

job charact. job charact. job charact. job charact. job charact. job charact.

earnings subj. chr. subj. chr. subj. chr. subj. chr.earnings optimism optimism

earningsObservations 3764 3743 3706 3068 3065 2550 3045 2536Pseudo_R2 0,0176 0,0365 0,0444 0,067 0,181 0,191 0,184 0,194Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

Page 21: Informality,  Welfare and Productivity

Table 3Probit-Mg Effects All Workers-Honduras 18-64 years oldDependent Variable: Job Satisfaction

(6)Salaried_small -0.0398*

(0.0229)Salaried_lg_nobenefits -0.0073

(0.0211)Salaried_lg_benefits 0.0617*

(0.0344)Hours (monthly) 0.0002**

(0.0001)Manufacturing -0.016

(0.0302)Const,retail,transport 0.0204

(0.0274)utilities, insurance and soc. serv. 0.011

(0.0278)Stressful -0.0183

(0.0142)Dangerous -0.0084

(0.0155)Monotonous -0.0558***

(0.0202)Progress Opportunity 0.0872***

(0.0170)Good Schedule 0.1188***

(0.0219)Well Remunerated 0.1280***

(0.0160)Insecure Job -0.0819***

(0.0211)Log Earnings (monthly) 0.0255***

(0.0080)Observations 2550Pseudo_R2 0.191

Omitted category self-employed

Page 22: Informality,  Welfare and Productivity

Table 3Probit-Mg Effects All Workers-Honduras 18-64 years oldDependent Variable: Job Satisfaction

(6)Salaried_small -0.0398*

(0.0229)Salaried_lg_nobenefits -0.0073

(0.0211)Salaried_lg_benefits 0.0617*

(0.0344)Hours (monthly) 0.0002**

(0.0001)Manufacturing -0.016

(0.0302)Const,retail,transport 0.0204

(0.0274)utilities, insurance and soc. serv. 0.011

(0.0278)Stressful -0.0183

(0.0142)Dangerous -0.0084

(0.0155)Monotonous -0.0558***

(0.0202)Progress Opportunity 0.0872***

(0.0170)Good Schedule 0.1188***

(0.0219)Well Remunerated 0.1280***

(0.0160)Insecure Job -0.0819***

(0.0211)Log Earnings (monthly) 0.0255***

(0.0080)Observations 2550Pseudo_R2 0.191

Omitted category self-employed

Earnings strongly associated withJob satisfaction

Page 23: Informality,  Welfare and Productivity

Table 3Probit-Mg Effects All Workers-Honduras 18-64 years oldDependent Variable: Job Satisfaction

(6)Salaried_small -0.0398*

(0.0229)Salaried_lg_nobenefits -0.0073

(0.0211)Salaried_lg_benefits 0.0617*

(0.0344)Hours (monthly) 0.0002**

(0.0001)Manufacturing -0.016

(0.0302)Const,retail,transport 0.0204

(0.0274)utilities, insurance and soc. serv. 0.011

(0.0278)Stressful -0.0183

(0.0142)Dangerous -0.0084

(0.0155)Monotonous -0.0558***

(0.0202)Progress Opportunity 0.0872***

(0.0170)Good Schedule 0.1188***

(0.0219)Well Remunerated 0.1280***

(0.0160)Insecure Job -0.0819***

(0.0211)Log Earnings (monthly) 0.0255***

(0.0080)Observations 2550Pseudo_R2 0.191

Omitted category self-employed

Job characteristics correlatewith job satisfaction in an Expected manner

Page 24: Informality,  Welfare and Productivity

Table 3Probit-Mg Effects All Workers-Honduras 18-64 years oldDependent Variable: Job Satisfaction

(6)Salaried_small -0.0398*

(0.0229)Salaried_lg_nobenefits -0.0073

(0.0211)Salaried_lg_benefits 0.0617*

(0.0344)Hours (monthly) 0.0002**

(0.0001)Manufacturing -0.016

(0.0302)Const,retail,transport 0.0204

(0.0274)utilities, insurance and soc. serv. 0.011

(0.0278)Stressful -0.0183

(0.0142)Dangerous -0.0084

(0.0155)Monotonous -0.0558***

(0.0202)Progress Opportunity 0.0872***

(0.0170)Good Schedule 0.1188***

(0.0219)Well Remunerated 0.1280***

(0.0160)Insecure Job -0.0819***

(0.0211)Log Earnings (monthly) 0.0255***

(0.0080)Observations 2550Pseudo_R2 0.191

Omitted category self-employed

Salaried small are worse off thanself-employed

Salaried at large firms arebetter off.

Page 25: Informality,  Welfare and Productivity

Results are robust to:

• Controlling for selection into employment• Results do not seem to be related to

unobservable variables correlated to job security and job category.

• Using a pseudo- fixed effect estimator to control for unobserved heterogeneity

Page 26: Informality,  Welfare and Productivity

Results change with level of education

Table 6Probit-Mg Effects by Education Level-Honduras 18-64 years old

No education level completed

Primary School Completed

High School Completed

College Completed

(1) (2) (3) (4)Salaried_small -0.0792* -0.0252 -0.0134 -0.0263

(0.0475) (0.0337) (0.0349) (0.0506)Salaried_large_nobenefits -0.0111 0.0059 0.0253 -0.0361

(0.0561) (0.0289) (0.0287) (0.0274)Salaried_large_benefits 0.0328 0.0376** -0.0167

(0.0615) (0.0180) (0.0426)Observations 754 1247 288 236Pseudo_R2 0.199 0.164 0.372 0.46Note 1: Controls in this regression are gender, age, age2, zone, levels of education. Marital status, hours worked,

health index, sector, and job attributes such as: stressful, dangerous, monotonous, progress opportunity, good schedule, well remunerated, insecure job, and optimism. Note 2: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

Page 27: Informality,  Welfare and Productivity

Results change across countries

Table 7Probit-Mg Effects All workers 18-64 years old

El Salvador Guatemala(1) (2)

Salaried_small -0.004 -0.2486***(0.0633) (0.0875)

Salaried_lg_nobenefits 0.0845* -0.1785***(0.0450) (0.0655)

Salaried_lg_benefits 0.057 0.0425(0.0565) (0.0693)

Observations 336 516Pseudo_R2 0.189 0.256Note 1: Controls in this regression are gender, age, age2, zone, levels of education. Marital status, hours worked,

health index, sector, and job attributes such as: stressful, dangerous, monotonous, progress opportunity, good

schedule, well remunerated, insecure job, and optimism.

Note 2: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

Strong preferenceFor self-employment

Page 28: Informality,  Welfare and Productivity

And education…

Table 8Probit-Mg Effects by Educacion- 18-64 years old. Other Countries.

Low Education High Education Low Education High Education(1) (2) (3) (4)

Salaried_small -0.1095 0.0425 -0.2053** -0.1463(0.1079) (0.0557) (0.0981) (0.0947)

Salaried_large_nobenefits -0.0811 0.0954* -0.1918** -0.0567(0.1364) (0.0488) (0.0772) (0.0546)

Salaried_large_benefits -0.1023 0.1210*** -0.0968 0.1371***(0.1732) (0.0437) (0.2391) (0.0463)

Observations 108 291 292 256Pseudo_R2 0.0139 0.0273 0.0359 0.0461Note 1: Low education = 1 if education 1 or education 2 are equal to 1; High education = 1 if education 3 or education 4 are equal to 1Note 2: No controls added.Note 3: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

Guatemala El Salvador

Higher preference for formal among skilled workers Lower valuation of benefits among Unskilled?

Page 29: Informality,  Welfare and Productivity

In sum

• Data suggests that unskilled workers have lower valuations of benefits.

– Lower financial/pension literacy – Cash constraints?

• They perceive payments as taxes; • Labor supply more elastic; taxes will tend to fall on

employers;• Which may explain why many unskilled workers are hired

without SS

• It may explain why so few workers blame the entrepreneurs for their lack of social security

Page 30: Informality,  Welfare and Productivity

Even if workers are better off in (some) informal jobs, informality leads to poor outcomes… • Large numbers of people possibly not

getting protected against poverty in old age.

• Do informal workers save for retirement?

• Informality renders labor policy ineffective.

• How to implement unemployment insurance, or protect workers against accidents or disability?

Page 31: Informality,  Welfare and Productivity

Part II: Informality lowers productivity

based on “Informality, Resource

Misallocation and Productivity in Brazil”

Carpio and Pagés (2009)

Page 32: Informality,  Welfare and Productivity

(A) channel: Informality leads to resource misallocation

• TFP (formal) > TFP (informal) firms. • Informal and formal firms produce goods

that are substitutes to some degree. • Informality is a subsidy on less productive

firms, allowing them to sell products to a lower price than they would if they paid taxes and regulations.

• This increses their market share• And reduces aggregate productivity.

Page 33: Informality,  Welfare and Productivity

Data

• Use a large and detailed sample of almost 39000 small firms in Brazil (less than 5 paid workers). Urban Informal economy 2003 (ECINF)

• It covers employers with less than 5 employees and self-employed workers, irrespective of the number of non-remunerated employees or partners.

• Definition of formality:

– Formal firms: Pay income tax– Informal firms: Do not pay income tax

Page 34: Informality,  Welfare and Productivity

We measure TFP and marginal product of capital and labor in each firm following Hsieh and Klenow (2009)

• Assumes CRS production function with shares computed from cost shares for each sector.

• If all firms in efficient allocation:

• MPK(i)=R• MPL(i)=w

• Departures from this indicate allocative problems.

Page 35: Informality,  Welfare and Productivity

Productivity (TFP) is higher in formal firms no matter how we measure it

Payroll: sum of wages for employees and self-employed without unpaid workers, excluding social security payments. We attribute wages to unpaid workers by using a Mincer regression. Capital: the value of capital stock at its market value.

0.1

.2.3

.4

kdensity logA

si

-6 -4 -2 0 2Log(Asi)

Registered Unregistered

Source: Based on Ecinf 2003

Mean difference55%

Page 36: Informality,  Welfare and Productivity

We measure the difference between MPL or MPK as a firm specific wedge.

Distortions that affect the price of Y relative to L

Distortions that affect the price of Y relative to L

Page 37: Informality,  Welfare and Productivity

Simple argument:

• If formal firms are characterized by high MPL and MPK it indicates that hiring more labor and more K would increase overall output and TFP.– Formal sector too small, Informal sector too large

• Instead, if MPL and MPK is higher in informal firms, it indicates they face constraints to growth. Productivity would increase if informal firms grew.

Page 38: Informality,  Welfare and Productivity

0

.2.4

.6.8

kdensity logT

FP

Rsi

-4 -2 0 2 4Log(TFPRsi)

Registered Unregistered

Source: Based on Ecinf 2003

On average, informal firms in Brazil are more constrained in their growththan formal firms.

In Brazil, high marginal products more likely among formal firms…

Page 39: Informality,  Welfare and Productivity

Bottom line…

• A larger number of formal firms should be larger than they are (they are taxed above average)

• A larger number of informal firms should be smaller than they are (they are relatively subsidized by not paying taxes)

• Since formal firms are more productive, the fact that they are too small reduces aggregate productivity.

• Other possible channels at play (less access to K & less innovation of informal firms)

Page 40: Informality,  Welfare and Productivity

A more proactive approach needed

• Creating social security and labor systems which are informality-proof (delinking them from the labor market, particularly for unskilled labor).

• Reducing the costs and increasing benefits of formalization

– Credit can help

Page 41: Informality,  Welfare and Productivity

Credit can help

• Catao and Pagés (2009).• Higher access to credit increases the

opportunity cost of informality• We explore an episode of increased supply

of credit in Brazil.• We find that formality rates (SS affiliation)

increased faster in sectors that are more credit dependent, and therefore benefit more from credit.

Page 42: Informality,  Welfare and Productivity

Conclusions

• Self-employed workers may not be worse off than salaried formal workers.

• Salaried workers in small firms tend to be less happy about their jobs.

• Higher preference for self-employment and lower for SS benefits for less skilled workers.

• Yet, informality even if optimal from an individual point of view is socially suboptimal

Page 43: Informality,  Welfare and Productivity

Conclusions

• Informality lowers productivity– How much?

• Informality renders social and labor policy ineffective

• State needs to be more proactive to deal with informality– Reducing costs of formality– Increasing benefits of being formal– Designing social security and labor policies

which are more informality-proof.

Page 44: Informality,  Welfare and Productivity
Page 45: Informality,  Welfare and Productivity

Ideas for when reviewing Madrigal Pages:

• Do fixed effect estimator by level of education

Page 46: Informality,  Welfare and Productivity

Ideas for when reviewing Carpio Pages:

• A few informal firms seem to have very high MPK and L.